CPK: the abbreviation of Complex Process Capability Index. It is an indicator used by modern enterprises to express process capabilities. Strong process capability can produce products with high quality and reliability. The process capability indicator is a method to express the process level. Its essence is to reflect the process qualification rate. The Study of Process Capability determines the extent to which these features meet specifications to ensure that the yield of finished products is above the required level and can serve as a basis for continuous improvement of the process. The specifications are divided into unilateral and bilateral specifications according to the upper and lower limits. Only the specification limit and specification center or only the specification limit and specification center are called Single-sided specifications. The upper and lower limits of specifications and the center value are available, while the upper and lower limits and the center value are symmetric specifications. When our product passes the gager & R test, we can start the CPK value test. The larger the CPK value, the better the quality. CPK-Process Capability Index CPK = min (cpku, cpkl) usl (upper specification limit): Maximum specification. LSL (low specification limit): minimum specification limit. Limit x = (X1 + X2 +... + Xn)/n: average value. T = usl-LSl: Capacity tolerances. U = (USL + LSL)/2: Specification center. Cpku = | usl-rjx |/3σ cpkl = | rjx-LSl |/3σ Significance of CPK Quantitative reflection of the Process Level (using a value to express the process level) process capability index is a convenient way to express the process level, the essence is to reflect the pass rate of the process. CPK Calculation Formula CPK = Cp * (| 1-ca |) Ca (capability of accuracy): Process Accuracy; CP (capability of precision): process precision; Note: When calculating CPK, the sampling data should contain at least 20 groups of data, which must be representative. CPK processing principles A + ≥ 1.67 no disadvantages consider reducing costs a1.33 ≤ CPK <1.67 well maintained status b1.0 ≤ CPK <1.33 improved to A-level c0.67 ≤ CPK <1.0 poor process many, dcpk must be improved. <0.67 the process capability is poor, and the design process should be improved. [1] CPK application description 1. CPK is defined in Chinese as: Process Capability Index, which is a quantitative reaction of a project or process level. It is also an indicator of engineering evaluation. 2. Two Parameters closely related to CPK: Ca, CP. Ca: Process accuracy. The consistency between the "actual average" and "Type center value" is measured. For a single-side specification, because there is no specification center, there is no Ca; for a bilateral specification, CA = (ˉ X-C)/(T/2 ). CP: process precision. Measure the ratio of "Capacity tolerance width" to "process variation width. For a single-side specification, only the upper limit and center value are available. CPU = | usl-0000x |/3σ. Only the lower limit and center value, CPL = | rjx-LSl |/3 σ for bilateral specification: Cp = (USL-LSL)/6 σ 3. the relationship among CPK, Ca, and CP: CPK = Cp * (1-| ca |). CPK is the neutralization reaction between Ca and CP, CA reflects the positional relationship (concentrated trend), and CP reflects the distributed relationship (discrete trend. when selecting the CPK for management and control, the cost should be the primary factor, and the impact of its quality characteristics on the subsequent process. 5. The calculated sampling data should contain at least 20 ~ 25 groups of data are representative. 6. Calculating the CPK should not only collect the sampling data, but also know the upper and lower limits of the specifications (USL, LSL) of the quality characteristics before the value can be smoothly calculated. 7. First, use the "stdevp" function of Excel.
(Note: it should still be "STDev". Refer to the data calculated by Minitab .) The standard deviation (σ) of the sampled data is automatically calculated, and the type tolerances (T) and the type center value (u) are calculated ). type tolerances T = upper limit of specifications-lower limit of specifications; type center values U = (upper limit of specifications + lower limit of specifications)/2; 8. according to the formula: CA = (X-U)/(T/2), calculate the Process Accuracy: Ca value (X is the average of all sampling data) Ca rating criteria and processing:
| Level |
CA Value |
Handling principles |
| A |
| Ca | ≤ 12.5% |
The operator must follow the operation standards and meet the requirements. |
| B |
12.5% ≤ | ca | ≤ 25% |
It is necessary to Grade. |
| C |
25% ≤ | ca | ≤ 50% |
The operator may see the error specification or fail to operate according to the operating standard. Specifications and operating standards shall be reviewed. |
| D |
50% ≤ | ca | |
Emergency measures should be taken to comprehensively review all factors that may affect production and, if necessary, stop production. |
9. Based on the formula CP = T/6 σ, the process precision is calculated: Cp Value CP rating criteria and processing:
| Level |
Cp Value |
Handling principles |
| A + |
CP ≥ 1.67 |
No disadvantage. You can consider reducing costs. |
| A |
1.33 ≤ CP ≤ 1.67 |
Maintain the status quo. |
| B |
1.00 ≤ CP ≤ 1.33 |
Improved to. |
| C |
0.67 ≤ CP ≤ 1.00 |
The process is poor, and the capability must be improved. |
| D |
CP ≤ 0.67 |
Due to poor process capability, the design process should be revised. |
10. according to the formula: CPK = CP (1-| ca |), calculate the Process Capability Index: CPK value 11. CPK rating criteria: (corresponding countermeasures can be taken to calculate the process capability index based on these criteria)
| Level |
CPK Value |
Handling principles |
| A ++ |
CPK ≥ 2.0 |
Cost reduction considerations |
| A + |
2.0> CPK ≥ 1.67 |
Excellent, should be maintained |
| A |
1.67> CPK ≥ 1.33 |
Good performance, good performance, and stable status, but should be upgraded to a + level |
| B |
1.33> CPK ≥ 1.0 |
Generally, a slight variation in process factors may lead to a negative risk. Various resources and methods should be used to promote it to a level. |
| C |
1.0> CPK ≥ 0.67 |
Poor, poor process, must improve its capabilities |
| D |
0.67> CPK |
Unacceptable, too weak, and a new design process should be considered. |
12. CPK and process yield conversion.
| CPK |
Poor performance of every 1 million pieces |
Pass rate |
| 0.33 |
317310 |
68.3 |
| 0.67 |
45500 |
95.5 |
| 1 |
2700 |
99.73 |
| 1.33 |
63 |
99.9937 |
| 1.67 |
0.57 |
99.99995 |
| 2 |
0.002 |
100 |
Both CPK and PPK represent the process capability parameters, and PPK adds attention to the special causes of the process, which is an indicator to describe the process performance. Minitab software is widely used in modern computing, which is convenient and convenient.
